Improvements in MR and Dynamic PET Imaging Algorithms

Arizona State University-Tempe Campus, Tempe AZ

Investigators

Linked publications & trials

Abstract

DESCRIPTION (provided by applicant): Reliable and highly sensitive neuroimaging PET/MR analysis tools for the assessment and detection of functional and anatomical changes in the human brain will be developed, implemented and statistically validated. The resulting novel and robust numerical algorithms will be fully integrated in a software package with a graphical user interface for greater utility, ease of use, and adoption by neuroimaging researchers. Particular emphases of the research design are a) development of completely automated procedures (including automated non-invasive determination of the input function) for the study of the cerebral metabolic rate for glucose (CMRglc) in tissue using the radioactive tracer 18F-fluorodeoxyglucose (FDG) with dynamic PET, b) high resolution reconstruction techniques for MR providing improved segmentation for more consistent detection of structural changes, and c) use of co-registered dynamic PET and high resolution reconstructed MR as an alternative approach for identification of FDG in plasma. Methodologies include a) the high resolution Gegenbauer reconstruction method with edge detection for MR, b) an efficient, improved clustering of 4D dynamic PET data by preclustering techniques for both automated extraction of the region of the interest of the Carotid Artery (for detection of tracer in plasma) and reduction in noise characteristics for generation of more accurate CMRglc and individual rate constant 3arametric images, and c) the Volterra integral formulation of the compartmental model of tracer dynamics for improved efficiency in numerical estimation of parametric data. Current tools will be enhanced and extended for development of a flexible software package providing an automated approach for neuroimaging studies by AIzheimer's dementia researchers at Banner Good Samaritan Medical Center in Phoenix. The statistically validated suite of tools will then be more widely available to other research groups, and enlarged for other tracer models for potential use in other disease studies of the brain.